A kinetic two-scale damage model for high-cycle fatigue simulation using multi-temporal Latin framework

Abstract : The goal of this paper is to introduce a model order reduction method for high-cycle fatigue simulations using a kinetic damage model, i.e. a constitutive model in which the damage evolution law is defined as a rate form for the damage variable D. In the framework of continuum mechanics, high-cycle fatigue simulation involves a two-scale damage model, which includes macroscopic elastic and microscopic plastic behaviours, for a very large number of cycles. Unlike the classical usage of the two-scale damage model by Lemaitre and co-workers, where damage is calculated as a post-process of an elastic or elasto-plastic macroscopic analysis, in this work, a fully coupled analysis is conducted assuming a macroscopic damage feedback from its microscopic counterpart. Damage is considered to be isotropic with micro-defect closure effect on both macroscopic and microscopic scales. To overcome the numerical expense, the large time increment (LATIN) method is used as a linearisation framework, where the constitutive behaviour is separated from the global admissibility which in turn is solved through separation of variables using a proper generalised decomposition (PGD)-based model reduction method. A multi-temporal discretisation approach is henceforth used based on finite element like description in time for the quantities of interest, providing a sophisticated numerical approach suitable for high-cycle fatigue simulation under complex loading.
Complete list of metadatas

Contributor : Amelie Fau <>
Submitted on : Tuesday, July 23, 2019 - 8:46:10 AM
Last modification on : Wednesday, August 21, 2019 - 3:49:27 PM



Mainak Bhattacharyya, Amélie Fau, Rodrigue Desmorat, Shadi Alameddin, David Néron, et al.. A kinetic two-scale damage model for high-cycle fatigue simulation using multi-temporal Latin framework. European Journal of Mechanics - A/Solids, Elsevier, 2019, 77, pp.103808. ⟨10.1016/j.euromechsol.2019.103808⟩. ⟨hal-02190831⟩



Record views